Dongsoo Lee

The paper analyzes how retinal ganglion cells adapt discriminability to natural scenes using a model of neural sensitivity and variability, finding that correlated noise reduces information transmission.

The paper classifies retinal amacrine cells by their responses to natural scenes using a neural network model and identifies seven clusters based on model presynaptic inputs.

The paper uses information geometry to analyze retinal representations of natural stimuli, finding that noise correlations limit information rather than enhance it, and that population coding improves information transmission.

The paper presents a mechanistically interpretable model of retinal ganglion cell responses to natural scenes, incorporating slow adaptive dynamics and capturing retinal computations.

The paper studies motion anticipation in retinal amacrine cells using optical and electrical recordings, finding that slow biphasic amacrine cells contribute most to motion anticipation.